The opening spends the only seconds that were guaranteed
Greeting, channel context and credentials all land before the reason to watch. By the time the useful part arrives, the audience that would have wanted it has already moved on.
Short-form video analyzerFind the line losing viewers
Find the line losing viewers
in your short-form video,
before you film it.
Paste your script and get your free score: one weighted score out of 100, five dimension scores behind it, and the exact line costing you the most — while changing it is still free.
No account, no card.
Scored forTikTokReelsShorts
Script check0 / 1,800
No script handy?
Example
41Weak · likely to lose viewers earlyBiggest issue: the point arrives too late
Hook 32Structure 48Pacing 31Rewatch 52Search 40
Costing you the most: “So basically, what I usually like to do first is…”
Free: your score, all 5 dimension scores and the line costing you the most. No account, no card.
In one paragraph
A short-form video analyzer scores how a TikTok, Reel or Short is likely to perform before it is published — by reading the script rather than the edit. Contextify's runs on text you paste, so you can improve your script before filming.
You get one weighted score out of 100, built from five dimensions: hook strength (30%), clarity and structure (25%), delivery and pacing (20%), rewatch and save (15%) and search optimization (10%). Every dimension score, its reasoning, and the line costing you the most are included. Your first rewrite is free with a free account. After that, 3 rewrites for $4.99, or Plus.
The scoring model
Five dimensions, and what each one is worth
Fixed, published weights, so two scripts are comparable. Not derived from your view data. Hook strength carries the most because nothing after it gets seen without it.
Hook strength30% of the score
The first three seconds decide the rest.
Clarity & structure25% of the score
One idea, delivered in order.
Delivery & pacing20% of the score
Filler and detours cause mid-video exits.
Rewatch & save15% of the score
Saves and shares outrank likes.
Search optimization10% of the score
Spoken words are indexed as search terms.
The diagnosis
Why good short-form videos underperform
Most underperforming short-form videos are not badly shot. They are well shot versions of a script that gives viewers permission to leave — and short-form platforms turn that permission into a distribution decision within the first few seconds.
TikTok, Reels and Shorts all rank on the same core signals: did people finish it, did they watch it twice, did they save it, did they send it to someone. All four are decided by what you say and the order you say it in. The edit can support a script that earns them. It cannot manufacture them.
Why the script is the part worth checking
Lighting, transitions and sound design are the expensive things to change and the least likely to be the reason a video stalled. A sentence is the cheapest thing in the entire production — and it is where nearly every avoidable retention problem lives.
That is the whole argument for checking before you post. After publishing, your analytics can only tell you that something went wrong. Before publishing, a script check tells you which sentence it was, while rewriting it is free.
Failure patterns
Five reasons scripts lose the room
These are the patterns the analyzer flags most often across the five dimensions.
The script branches instead of building
Three loosely related tips read as generous. To a viewer they read as a list with no destination — there is no single payoff to wait for, so leaving at any point costs nothing.
Filler stretches a good idea past its welcome
“So basically, what I usually like to do…” delays the instruction by two full seconds. Repeated four or five times across a script, that is most of your retention curve.
Nothing in it is worth saving or sending
Saves and shares are the highest-intent signals short-form platforms collect, and they are weighted well above likes. A script with no reusable line, number or steal-able phrase never earns either.
The words nobody would search for
Spoken words are transcribed and indexed. A script that never says the term its audience would type is invisible to the search surface sitting next to the feed.
Scored examples
The same point, two scripts
Nothing here changes the idea. What changes is where the useful part sits in the sentence.
Buries the point
“So basically, what I usually like to do first is start by writing out a list of everything I need to get done…”
Pacing 31 / 100Arrives on time
“Before you open your inbox, write down the one task that actually matters today.”
Pacing 86 / 100Seven setup words removed. The instruction is now the first thing the sentence delivers.Branches outward
“I also use a timer, drink more water, and sometimes I switch to a different app…”
Structure 38 / 100Builds to one payoff
“Try it tomorrow: one task, twenty focused minutes, then you are allowed to open your inbox.”
Structure 82 / 100One idea carried to a specific instruction — which is also the line most likely to be saved.Illustrative examples. Your own scores are generated from your words.
Three steps
How to analyze a short-form script
Two minutes end to end, and no footage required to find out.
Paste the whole script, not just the opening
Everything you say from the first word to the last. Already filmed it? Paste the captions or your teleprompter text. It runs on text, so you can check a script before you commit a take to it.
Sized for 30–60 second posts — roughly 40 to 150 spoken wordsRead one weighted score and five dimension scores
Hook strength, clarity and structure, delivery and pacing, rewatch and save, and search optimization are each scored 0–100 and combined using published weights. Every dimension comes back with a sentence describing what is happening in your script, so you can see which one is dragging the total down.
The weights are fixed, so two scripts are directly comparableSee the line costing you the most
The analyzer quotes the exact passage doing the most damage and explains what it costs you across the dimensions it touches. The specific change and a rewritten version are the part you unlock. Your first rewrite is free with a free account. After that, 3 rewrites for $4.99, or Plus.
Your score is kept for seven days, account or not FAQ
Check a script instead →Questions about script scoring
What is a short-form video analyzer?
A short-form video analyzer scores how a TikTok, Reel or Short is likely to perform before it is published, by evaluating the script rather than the edit. Contextify's returns one weighted score out of 100 plus five dimension scores — hook strength, clarity and structure, delivery and pacing, rewatch and save, and search optimization — with the exact line costing you the most quoted back to you.
Is the script analyzer free?
Yes, the check is free with no account and no card: you get your overall score, all five dimension scores and the exact line costing you the most, with what it costs you. Anonymous checks are limited to 3 per hour. Your first rewrite of that line is free with a free account. After that, 3 rewrites for $4.99, or Plus. New members can try Plus for 7 days with no card; to keep it, Plus is $14.99 CAD/month.
How are the five dimensions weighted?
Hook strength counts for 30% of the total, clarity and structure 25%, delivery and pacing 20%, rewatch and save 15%, and search optimization 10%. Hook carries the most weight because nothing downstream matters if viewers leave in the first seconds. The weights are fixed and published, so two scripts scored weeks apart are directly comparable.
How is this different from checking my analytics?
Analytics describe what already happened. By the time a completion rate shows up in your dashboard, the post has been distributed and the result is fixed. A script analyzer runs before publishing, while changing a sentence is still free.
Do I need to upload a video?
Not for a script check — it runs on pasted text, so you can test a script before recording it. Uploading a real video, where the audio is transcribed automatically and scored in the same five dimensions, happens inside your account.
What platforms does this cover?
TikTok, Instagram Reels and YouTube Shorts. All three rank primarily on completion, saves and shares, which is what the five dimensions are built around. The scoring does not change between them, so one check applies to a post going out on all three.
What does the overall score actually mean?
It is the weighted combination of the five dimension scores, not a separate judgment. Below 25 is critical, 25 to 49 is weak, 50 to 74 is workable, and 75 and above is strong. Because hook strength carries 30%, a weak opening pulls the total down even when everything after it is solid — which is deliberate.
Is my script stored or used for training?
Your check is encrypted and kept for seven days so it survives a reload or a new tab, then deleted. With an account it is saved to that account instead. Scripts are never published, never shared, and never used to train models.
Does a high score guarantee views?
No. The analyzer measures the script — the part you control before posting. It cannot account for your topic's demand, your audience, your timing, your edit or your luck. Use it to remove the avoidable problems, not to predict a number.
Find out which line is costing you the room
Paste the script. Get the weighted score, all five dimensions, and the passage doing the most damage — before you commit a take to it.
Free score, no account, no card. Plus: a 7-day trial with no card, then $14.99 CAD/month if you keep it. Keep going